This is the mail archive of the java-discuss@sourceware.cygnus.com mailing list for the Java project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: _Jv_HashCode


Tom Tromey wrote:
> 
> >>>>> "Jeff" == Jeff Sturm <jeff.sturm@appnet.com> writes:
> 
> Jeff> It seems to me that synchronization objects could provide their
> Jeff> own hash algorithm as well.
> 
> What could it be except object address?  It has to be something that
> never changes, regardless of the object's contents.  Also, we don't
> want to add an extra word to the object, because getting rid of a word
> is one point of making this change.

Like you said, the bottom 2 or 3 bits of the address are useless, so you
can shift right.

Perhaps _Jv_HashCode should be changed to do the same after all, if gdb
can be fixed.  On 64-bit archs that could shift in some meaningful bits
too.

-- 
Jeff Sturm
jeff.sturm@appnet.com

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]